Slide-Level Active Learning Reduces Annotation Burden in H&E images

2026-07-10

AI Summary

This work addresses the high annotation cost in whole-slide H&E image segmentation and limitations of existing active learning methods in uncertainty estimation, annotation granularity, and class imbalance. The authors propose SHAL, a novel slide-level hybrid active learning framework that integrates predictive entropy and epistemic uncertainty, guided by foreground masks, class-priority sampling, and a multi-stage adaptive selection strategy to explicitly mitigate background interference and class imbalance. Evaluated on TCGA colorectal cancer data, SHAL achieves Dice scores ≥0.80 using only 26% of the annotation budget, demonstrates superior external generalization performance, and exhibits the smallest internal–external Dice gap (approximately 0.025), significantly enhancing both annotation efficiency and model robustness.

Abstract

Deep learning-based segmentation of histopathology whole-slide images (WSIs) requires large amounts of pixel-level annotations, which are costly and time-consuming to obtain. Active learning (AL) has been proposed to reduce this effort, but existing methods exhibit three key limitations. Uncertainty estimation is unreliable on partially annotated WSIs, patch-level acquisition is inconsistent with slide-level annotation workflows, and class imbalance in multi-class settings is not explicitly addressed. To address these challenges, we propose SHAL (Slide-level Hybrid Active Learning), a patient-level AL framework for annotation-efficient multi-class histopathology segmentation. SHAL integrates three complementary components: a foreground-aware strategy that suppresses bias from unlabeled background regions, a stage-adaptive mechanism that hybridizes predictive entropy and epistemic uncertainty across learning stages, and a class-aware strategy that prioritizes diagnostically relevant tissue classes. SHAL is evaluated on the TCGA colorectal cancer dataset. It achieves the highest Macro Dice at the full annotation budget (0.846) and reaches Dice greater than or equal to 0.80 using only 26 percent of the budget (50 of 190 slides), whereas competing methods reach this threshold only at 37 percent (70 slides). Across five independent external cohorts, SHAL attains the highest mean external Macro Dice (0.815) and the smallest internal-to-external generalization gap among all methods (0.025 at Round 3 and 0.026 at the full budget). The results indicate that patient-level hybrid uncertainty acquisition reduces annotation cost without sacrificing cross-domain generalization in computational pathology.

Problem

• active learning
• whole-slide image
• annotation burden
• class imbalance
• histopathology segmentation

Innovation

• slide-level active learning
• hybrid uncertainty
• foreground-aware
• class-aware sampling
• computational pathology